Demand Forecasting

Demand forecasting uses historical and external signals to estimate future material needs and support inventory, production, and logistics planning.

Business impact

• Profit per mill hour — The CEO says the company applied AI to improve profit per mill hour.
• Hot rolled steel produced — The article reports over 63,000 tons in the first full month of production, but it does not state that this was caused by the forecasting system.

Data requirements

• Macroeconomic data (tabular) — Used as an input to demand prediction models.
• Historical demand for steel (tabular) — Used as an input to machine learning models.
• Manufacturing activity data (tabular) — Used as an input to demand prediction models.
• Activity of large consumers of steel (tabular) — Examples in the article include housing starts and oil rig counts.
• Sensor data (sensor) — The platform is described as leveraging rich sensor data from the mill.

AI methods and techniques

• Machine learning — Used to build demand prediction models from macroeconomic and historical data.
• Predictive analytics — Applied through predictive AI engines to estimate future demand and support planning.

Industries

Real-world evidence

11 documented case studies on record.

Companies using this: Amazon, American Eagle, GE Aerospace, General Mills, Home Depot, Kroger, Merck, P&G, Renault, Schneider Electric, Big River Steel.